Knowledge Representation using a 3D GUI

[MuFu]

The time has come for me to choose a final year project, and although a couple of the suggested topics look quite interesting I'm considering writing my own criteria.

We've seen the introduction of VRML, Java3D etc pass relatively unnoticed, but now with Longhorn on the horizon and a huge web services push there is more opportunity that ever for apps to take advantage of 3D interfaces for knowledge representation and data visualisation.

Do you think the topic is one that's worth pursuing? I was thinking of tying in a substantial research effort (in the area of HCI, IKM etc) with some app development using XAML or 3D XML interfaces. Aside from the more obvious use of 3D for graphing/realisation, I'm interested in looking at how project management and other business applications can benefit from the extra dimension. I'm just quite apprehensive because it seems like an area that would have received a lot more attention by now, had it actually offered more in the way of practical utility.

The project is worth 30 credits which corresponds to about 300 hours of work. Incidentally the other topics I'm considering are "Texture Synthesis by Non-parametric Sampling" and "Digital Musician - an application to read musical scores, recognise them and produce digital audio relevant to the printed notes".

 

[Frank]

I would suggest you check out all the things they did at Palo Alto, the Xerox research lab. It provides you with any amount of yummy things to research, including multiple ways to represent data in 3D.

 

[Zaphod]

"Digital Musician - an application to read musical scores, recognise them and produce digital audio relevant to the printed notes".

Do you mean an OCR solution for sheet music? Several such programs already exist.

 

[MuFu]

I would suggest you check out all the things they did at Palo Alto, the Xerox research lab. It provides you with any amount of yummy things to research, including multiple ways to represent data in 3D.

Thanks, that looks like a useful resource.

"Digital Musician - an application to read musical scores, recognise them and produce digital audio relevant to the printed notes".

Do you mean an OCR solution for sheet music? Several such programs already exist.

Yep. I know of SharpEye but haven't looked into other solutions yet. Obviously the challenge here would be to produce something even better - that can read handwritten scores perhaps?

 

[Zaphod]

Yep. I know of SharpEye but haven't looked into other solutions yet.

There's also PotoScore and SmartScore (at least). Saw one of them demoed a while ago (can't remember which). Very impressive.

Obviously the challenge here would be to produce something even better - that can read handwritten scores perhaps?

Kind of ambitious for an eight week project, no? Good luck! Musicians everywhere would love you if you could pull that one off.

 

[MuFu]

Cheers. Yep - I would have thought that'd be far too ambitious, although it might be a continuation of work that's already taken place here at uni by PhD students, in which case there should be quite a bit of relevant image processing/pattern recognition knowledge available. Still, probably a no-go given the constraints (and my complete lack of experience in OCR, hahaha).